SYNTHESIS AND MASSSPECTRA OF SUBSTITUTED QUINAZOLINONES J . Org. Chem., Vol. 37, No. ence between a methiodide and a corresponding protonated system is an axial methyl substituent then an axial methyl @-substituent parameter at C-2,6 can be derived. I n the 4-piperidones an average downfield shift of 7.1 ppm is derived while in the piperidines the corresponding value is 8.1 ppm. The CH8NCH3,,+ parameter given in Table I11 for the piperidines (10.5 ppm) was derived from the data on the methiodides of 1,2-dimethyl- (VIII) and 1,2,6-trimethylpiperidine (IX) and consequently comprises an effect from the Nmethyl-axial substituent (8.1 ppm) and the C-2’ methyl effect for the hydrochlorides (2.2 ppm). The sum of these values (10.3 ppm) is remarkably close to the composite value (10.5 ppm) as given in Table 111. Similarly, in the piperidones the corresponding value (4.6 ppm) comprises the axial @-substituent effect (7.1 ppm) and the “charge effect” a t C-2,6 (-3.1 ppm) providing the sum 4.0 ppm, also in close agreement with the composite value. These summations add further justification to the parameterization procedure suggested by the results of the present work. The effects of methiodation a t the C-3,s and C-4 sites follow similar trends to the hydrochlorides, though their magnitudes are greater. Further effects are also apparent at the substituent methyl carbons. The N-methyl carbons are found downfield 11.2 and 2.1 ppm, on the average, compared with the N-methyl carbon in the neutral molecule, while the shift a t C-2’ in the 2-methyl-4piperidones is shifted upfield an average of 4.6 ppm on methiodation. This latter shift is presumably steric in origin (see ref 35). A similar shift effect at C-3’ in the hydrochlorides (0.8 ppm) of the 3-methyl-4-piperidones VI and VI1 is also observed in the methiodides (0.9 ppm) and is also attributed to the steric effects. The methiodides of substituted 4-piperidones are generally hydrolyzed in aqueous solution.9~10 The
14, 1976
2337
carbon-13 shifts of the resulting geminal diols are given in Table I and although the upfield shift (approximately 100 ppm) at C-4 is clearly indicative of the change from a carbonyl to a geminal diol center no trends are clear for the other ring carbons. Conclusions The substituent additivity parameters summarized in Table 111, in addition to the considerations outlined in the Discussion section, should provide a useful basis for the analysis of structure and conformation in more complex molecules. I n subsequent studies, to be published, we have used these parameters with some success in order to determine conformations in a series of synthetic analgesic^.^^ Registry No.-IV, 1445-73-4; IV HC1, 34737-83-2 ; IV CHJ, 34737-84-3; V, 13669-32-4; V HC1, 3473786-5; V CH31, 34737-87-6; VIa, 4629-80-5; VIa HC1, 4629-76-9; VIa CH31, 34737-88-7; VIb, 34737-89-8; VIb HC1, 26822-34-4; VIb CH31, 34737-91-2; VII, 20281-02-1 ; VI1 HCI, 29849-51-2; VI1 CHJ, 3473792-3; VIIIa, 109-05-7; VIIIb, 671-36-3; VIIIb HC1, 5072-43-5; VIIIb CHJ, 34737-95-6; IXa, 504-03-0; IXb, 669-81-8; IXb HC1, 5072-29-7; IXb CHJ, 34737-99-0. Acknowledgments. -This work was supported in part by the National Research Council of Canada, Grant A6416. We thank the referee for some helpful comments and Mr. Glen Bigam for spectrometer maintenance. (36) E.Q., A. J. Jones, A. F. Casy, and K. M. J. MoErlane, Tetrahedron Lett., 1727 (1972).
Synthesis and Mass Spectra of Some 2-Carbomethoxy-2-carbomethoxymethyl-2,3-dihydro-4( 1H)-quinazolinones THOMAS F. LEMKE,*~ HARRYW. SNADY,JR.,AND NEDD. HEINDEL Department of Chemistry, Lehigh University, Bethlehem, Pennsylvania 18016 Received November 25. 1971
A synthesis of 2-carbomethoxy-2-carbomethoxymethyl-2,3-dihydro-4(1H)-quinazolinones is described and the major electron impact fragmentation pathways of this series are discussed.
We have recently noted that enamines 1, prepared by condensation of anthranilamides and dimethyl acetylenedicarboxylate, are versatile precursors of several classes of heterocyclics.2 Treatment of 1 with NaOXe in xylene produces a new class of benzodiazepinediones 4,which are labile in alcohol. The reaction of 1 with NaORIe in RIeOH leads directly to maleimides 2 and quinazolinones 3, which can also be obtained in the same ratio in a separate experiment as the ring contraction products of the benzodiazepinediones, thus pointing toward 4 as a possible intermediate. (1) Department of Chemistry, Marshall University, Huntington, 71;. Va. 25701. (2) N. D. Heindel, V. B. Fish, and T. F. Lemke, J . 070. Chenz., 38, 3997 (1968).
The isolation of pure quinazolinones by this technique is tedious, since careful fractional separation from the maleimide coproducts is necessary, and only three quinazolinones (3a-c) were prepared by this route. With an N-methyl group on either the amide or amino nitrogen of the anthranilamide, steric hindrance apparently prevents formation of a benzodiazepinedione intermediate and only quinazolinone (3d or 3e) (not admixed with maleimide) results in base-catalyzed cyclization of the corresponding adduct. We have now found that the quinazolinones 3 can be obtained uncontaminated by benzodiazepine or maleimide coproducts, by simple thermolysis of the enamino adducts 1. The quinazolinones obtained displayed two saturated ester carbonyls in their ir spectra between 1725 and 1750 cm-l and a characteristic CH2resonance
2338 J . Org. Chem., Vol. 37, No. 14, 1972
LEMKE,SNADY, AND HEINDEL TABLE I MASS SPECTRA OF 2-CARBOMETHYL-2-CARBOMETHOXYMETHYL3-R1-2,3-DIHYDRO-6-R~-4(H)-QUINAZOLINONESa Molecular ion,
4
4
m/e
3a
278
3b
312
3c
292
3d
292
3e
306
3f
388
3g
326
3 Ri
R2
Rs
a H H b H H C H CH3 d CH3 H e CH3CH3 f H CGH, g CH3 H h H CH3 i H H j H H
H C1 H H H C1 C1 C1 OCH3
F
R, H
H H H H H H H C1 H
in the nmr at 6 3.18 rt 0.10 ppm. The mass spectral fragmentation patterns for several of the quinazolinones were tabulated (Table I) for the purpose of assigning fundamental pathways. Very few electron impact studies have been performed on this heterocyclic clam3 The molecular ions were observed for all the quinazolinones at very low abundancies (less than 0.1% of the base peak) except for the dimethyl compound 3e, which has a parent ion of 2.7 relative intensity. The nature of the substituents at C-2 in all the quinazolinones was clearly demonstrated by the appearance of the generally abundant ions, a and b, resulting from
ir
9
m / e (re1 abundance)
219 (loo), 205 (25), 187 (as), 173 (4.4), 160 (24),159 (96),146 (lo), 145 (31), 132 (12), 130 (15),129 (12),120 (8.8), 119 (13),91 (14),90 (7.0),89 (14) 253 (14),238 (75), 221 (LO),193 (7.6), 180 (loo),179 (27),178 (32),160 (12), 153 (SO), 124 (421,74 (35) 233 (loo),219 (14), 201 (15), 173 (35), 159 (15),146 (E), 132 (20),104 (7.7) 233 (loo), 219 (23),202 (7.9),201 (49), 174 (8.8), 159 (8.8), 134(lo), 132 (7.9), 105 (16),104 (16),78 (lo), 77 (18) 306 (2.7),248 (15),247 (loo), 233 (12), 215 (5.4),188 (5.4),160(l5),145 (8.0), 114 (15),73 (14) 329 (53), 328 (loo), 315 (14),297 269 (57),255 (9),234 (8))180 (8), 153 (9), 132 (54), 125 (7),124 (lo), 117 (IO), 91 (9),77 (40) 326 (2.5),267 (loo), 253 (11), 235 (33), 208 ( L O ) , 193 (6.3),168(ll), 166 (4.4), 139 (9.4),138 (9.4),111 (6.3),77 (8.8), 75 (7.5)
a Values of ion abundancies (relative to base peak) of less than 1% and of ions containing the minor isotope of chlorine are omitted from the table.
process is envisioned as proceeding through a concerted, cyclic intermediate leading to species c. That the overall transition to these P - 91 ions did in fact encompass a distinct 32 mass loss was indicated by the presence of the appropriate metastable ions (Table 11). TABLE I1 METASTABLE IONS Compd
3a
0
0
3b 3c 3d 3e 3g
m*
Transition
159.5 135.4 104.5 130.0 173.5 173.5 187.5 207
219 + 187 187 + 159 159 + 129 180 4 152 233 + 201 233 + 201 247 + 215 267 + 236
Calcd
159.7 135.2 104.7 130.1 173.4 173.4 187.1 206.8
b
a
a “type AI” cleavage4 of carbomethoxy (.CO&Ie, 59 mass units) and carbomethoxymethyl ( -CH2CO2N!e,73 mass units), respectively, from the molecular ion. Another general feature of the cracking patterns was the loss of methanol (mass 32) from the a species. The
0
Compd
This same pathway was not available to the dimethyl derivative 3e, and thus an alternative route t o the m/e 215 ion appears to be the loss of methanol totally from the side chain, leading to species d.
0
d
( 3 ) G. Spiteller, Aduan. Heterocycl. Chem., 7 , 343 (1966). (4) K. Biemann, “Mass Spectrometry: Organic Chemical Applications,” McGraw-Hill, Kern York, N . Y., 1962, pp 76-77.
The P - 91 species ( e or d) further fragmented with losses of either CO (mass 28) or ketene (mass 42). The product of the first loss is the result of losing the side
SYNTHESIS AND MASSSPECTRA OF SUBSTITUTED QUINAZOLINONES J. Org. Chem., Vol. 37, N o . 14, 1972 2339 chain carbonyl function to form e or its ring-expanded version f ,
f
e
This transition was verified by the occurrence of a metastable peak for 3a at 135.4 (Table 11). This process did not occur for any of the quinazolinones with a methyl group at Tu'-1. Support for the intermediacy of a ring-expanded species was evidenced in at least one case by the facile loss of CH2NH2(mass 30) from 3a (metastable ion at 104.5),plausible only for specics f. Formation of ion g was observed for all the quinazolinones. This ion could arise by a McLafferty rearrangement of a or by loss of ketene from c. 0
0
H h
I n the cases of 3a,b this species could also have arisen from the loss of a hydrogen from h, which was produced by consecutive losses of .C02Me and CH2CO2nk from the molecular ion. Other prominent features of the high-mass portion of the spectra were the large abundances of the m/e 238 ion for 3b and the m/e 325 ion for 3f arising from McLafferty rearrangements with concomitant losses of CH3C02Me and HC02i\Ie, respectively, from the molecular ions. Some of the compounds showed a tendency to undergo a retro Diels-Alder cleavage (type D)&in the same manner reported by Budzjkiewicz, et CZZ.,~ for a series of quinazolinones in which the preferred cleavage of atoms 2 and 3 led t o species i.
-
it
1
Important fragmentations of this type were observed for 3a and 3f, leading to ionsmle 119 (p - 159) and 153 (p - 236), respectively. The contribution of this fragmentation pattern (3b) (m/e 153, p - 159) was uncertain due t o the ease at which another process occurred, verified by the presence of a metastable ion at m/e 130 and due to the loss of HCN (mass 27) from the base peak. (5) F. %'. McLafferty, "Mass Spectrometry of Organic Ions." Academic Press, New York, N. Y., 1963, p 423. (6) S.C. Pakrashi, J. Bhattaoharyyrt. L. F. Johnson, and H. Budeikiewioz, Tetrahedron, 19, 1011 (1963).
Experimental Section Melting points were determined between glass slide covers on a Fisher-Johns block and are reported uncorrected. Nmr spectra were determined on a Varian A-60 spectrometer using TMS as an internal standard. Mass spectra were run on a Hitachi Perkin-Elmer RMU-6E high-resolution instrument equipped with a double-focusing sector and a direct solids inlet system. All of the compounds were analyzed by placing approximately 1 mg of the sample into a 1-mm i.d. quartz tube which was introduced into the ion source of the spectrometer by a vacuum interlock. It was necessary to elevate the temperature of the direct inlet system to 80" in order to achieve ionization a t 80 eV. At this temperature, a quite singular mass spectrum was obtained for each of the compounds. High mass calibration of the spectra was accomplished by the use of tungsten hexacarbonyl as an internal standard. Microanalyses were performed by the late Dr. V. B. Fish of these laboratories and by Dr. G. I. Robertson of Florham Park, N. J. Anthrani1amides.-These materials, with exceptions as noted, were prepared by nucleophilic ring opening of isatoic anhydrides with amines and have been described in the literature.2~7~8 Preparation of Dimethyl Acetylenedicarboxylate Adducts of Anthrani1amides.-Reaction of equimolar amounts of the anthranilamide and dimethyl acetylenedicarboxylate in refluxing MeOH (0.1 mo1/100 ml of solvent) produced the desired adducts. This procedure has been previously described for l a 4 and for li-js and was applied herein to the preparation of Id-h. The adduct l e , prepared from 2-(N-methylamino)-Nmethyl benzamide, was a viscous oil and was utilized directly in synthesis of the quinazolinone. Id was prepared in 71% yield by reaction of 2-(Ar-methylamino)benzamide and dimethyl acetylenedicarboxylate. Light yellow crystals, mp 161-5-163" (from benzene), were obtained. Anal. Calcd for C1bHlBN206: C, 57.52; H , 5.51; N , 9.58. Found: C, 57.81; H, 5.77; N,9.74. If, prepared by condensation of 2-amino-5-chlorobenzanilide and the alkyne ester, was obtained in 71% yield, mp 143-145" (from MeOH). Anal. Calcd for C1~H&1N2O;: C, 58.69; H , 4.41; N , 7.20. Found: C, 58.65; H,4.21; N, 7.40. l g was obtained by reaction of 2-(N-methylamino)-5-chlorobenzamide with dimethyl acetylenedicarboxylate in 80% yield. The analytical sample, mp 171.5-172.5", was isolated by recrystallization from MeOH. Anal. Calcd for Cl4Hl6C1N2Oj: C, 51.46; H , 4.63; N , 8.57. Found: C, 51.72; H,4.48; N,8.43. lh, prepared in 7570 yield by condensation of 2-amino-5chloro-AJ-methylbenzamide and dimethyl acetylenedicarboxylate, melted a t 163-165' after two recrystallizations from RleOH. Anal. Calcd for C1dHljClN206: C, 51.46; H , 4.63; N, 8.57. Found: C,51.51; H,4.60; N,8.28. l e was obtained as a viscous oil by condensation of 2-(N-methy1amino)-N-methylbenzamideg with acetylenedicarboxylate by the previously described method.2,8 This oil was converted to the quinaxolinone 3e by redissolving in MeOH containing 0.1 g of NaOMe, then refluxing for 24 hr, concentrating, and diluting with water. The crude product precipitate (63%) was recrystallized from 95: 5 cyclohexane-benzene to yield the analytical material, mp 85-86.5". Anal. Calcd for C16H18N205:C, 58.82; II, 5.92; N, 9.15. Found: C, 58.72; H,5.81; N, 9.24. Thermal Cyclizations to Quinazo1inones.-The other quinazolinones reported herein were prepared by fusion of the anthranilamide adducts at 15-30' above their melting points. A sample procedure, illustrated by the synthesis of 3f, follows. 6-Chloro-2-carbomethoxy-2-carbome thoxymethyl-2,3-dihydro3-phenyl-4-(1H)-quinazolinone(3f).-A 1.94-g (5.0 mmol) sample of If was heated in an open tube at 165-175' for 2 hr. The cooled residue was then triturated with a mixture of equal parts of benzene and cyclohexane and filtered. The collected crystals (1.53 g, 797,) were washed with cyclohexane and re(7) N. D. Heindel, W. P . Fives, T. F. Lemke, and R . A. Carrano, J . Pharm. Sei., 60, 703 (1971). ( 8 ) N. D. Heindel, W.P. Fives, T. F. Lemke, J. E. Rowe, H. W. Snady, and R . A. Carrano, J . Med. Chem., 14, 1233 (1971). (9) H. Yale, J . Heterocycl. Chem., 8 , 193 (1971).
2340
J . Org. Chem., Vol. $7, No. 14, 1972
WEHRLI,FRYER, PXGOTT, AND SILVERMAN
crystallized from methanol to yield the pure quinazolinone 3f: mp 19S-2W0; ir (Nujol mull) 3260 (NH), 1752 and 1738 (ester CO), and 163s cm-l (amide CO); nmr (DMSO-de)6 7.93 ( s , 1, NH, exchanges with DzO), 7.38-7.26 (m, 7, ArH), 6.85 (d, 1, J O = S.5 Hz, ArHs), 3.64 ( 8 , 3, OCHB),3.19 (s, 3, OCHa), and 3.13 ppni (s, 2, CHZCO). Anal. Calcd for C19H17C1N205: C, 58.69; H, 4.41; N, 7.20. Found: C, 38.94; H, 4.20; N, 7.19. 2-Carbomethoxy-2-carbomethoxymethyl-6-chloro-2,3-dihydro1-methyl-4(lH)-quinazolinone (3g).-Pyrolysis of l g provided an 88% yield of 3g, which was recrystallized from benzene, mp 1S0-1S2". Anal. Calcd for C1dH1;ClNlOa: C, 51.46; H, 4.63; N, 8.57. Found: C, 51.64; H , 4.53; N, 5.67. 2-Carbomethoxy-2-carbomethoxyme thyl-6-chloro-2,3-dihydro3-methyl-4( lH)-quinazolinone (3h).-A 947, yield of 3h was obtained from Ih. Recrystallization from 1: 1: 1 MeOHbenzene-cyclohexane provided the analytical sample, mp 147.5149'. Anal. Calcd for C14Hl&lNZOb: C, 51.46; H, 4.62; N , 5.57. Found: C, 51.53; H,4.47; N,8.67. 2-Carbomethoxy-2-carbomethoxyme thyl-6-me thoxy-7-chloro2,3-dihydro-4(1H)-quinazolinone (3i).-3i was obtained by pyrolysis of li. Four recrystallizations from MeOH gave the purified product (36%),1° mp 169.5-171.5'. (10) The low yields of these compounds are partially a result of the increased solubility in the recrystallization solvents. K O attempt was made t o optimize the yields.
Anal. Calcd for C1,H&lNnOe: C, 49.06; H, 4.41; N, 8.17. Found: C, 49.36; H, 4.58; N, 8.07. 2-Carbomethoxy-2-carbomethoxymethyl-6-fluoro-2 ,J-dihydro4( 1H)-quinazolinone (3j).-This compound was prepared in 36%"0 yield from lj, mp 178-180' (after recrystallization from MeOH). Anal. Calcd for C I B H ~ ~ F N ~C,O52.70: ~: H, 4.42; N, 9.45. Found: C, 52.92; HI 4.48; N, 9.42. 2-Carbomethoxy-2-carbomethoxymethyl-2,3-dihydro-l-methyl4( 1H)-quinazolinone (3d).-3d was prepared by dissolving 3.0 g (0.01 mol) of Id in 50 ml of xylene containing 0.1 g of NaOhIe and heating to reflux for 3.5 hr. Chilling precipitated 1.32 g of unreacted Id. Dilution of the filtrate with petroleum ether (bp 30-60") gave 1.18 g of product. Two recrystallizations from MeOH yielded 1.02 g (34% conversion) of pure 3d,1° mp 118120". Anal. Calcd for ClaHleNzOj: C, 57.52; H, 5.51; N, 9.58. Found: C, 57.73; H, 5.41; N,9.53.
Registry No.--Id, 34804-41-6; l e , 34804-42-7; If, 34504-43-8; lg, 34804-44-9; lh, 34804-45-0; 3a, 17244-35-8; 3b, 17244-36-9; 3 ~ 717244-40-5; 3d, 34504-49-4; 3e, 34804-50-7; 3f, 34504-51-5; 3g, 3480390-2; 3h, 34803-91-3; 3i, 34803-92-4; 3j, 34803-93-5. Acknowledgment. -We gratefully acknowledge the financial support of the doctoral research of Thomas F. Lemke by Atlas Chemical Industries.
Synthesis of Trimethylhydroquinone from Aliphatic Precursors PIUSA. WEHRLI,*~ R. IANFRYER, F. PIGOTT, AND GLADYS SILVERMAN Chemical Research Department, Hoffmann-La Roche Inc., ATutley,New Jersey 07110 Received January 66, 1976 Condensation of a properly substituted a,p-unsaturated aldehyde or ketone with an appropriate aliphatic ketone leads to a trimethyl-2-cyclohexen-I-onestructure. These molecules are readily aromatized to the corresponding phenols in high purity with Pd/C under mild conditions. A novel indirect electrolytic oxidation involving Fremy's radical2 yielded trimethyl-p-benzoquinone (9) which in turn was reduced to trimethylhydroquinone (1). The latter is an important intermediate in the synthesis of vitamin E.
Various methyl-substituted phenols and particularly trimethylphenols are prepared from phenol via methylation processes. Although these processes can be controlled so that a particular mono- or polymethylated phenol can be obtained as the major product, mixtures are invariably obtained. Such mixtures require separation by methods which can be tedious especially if phenols of high purity are required. We now wish to describe a general method for obtaining a number of specifically substituked trimethylphenols in pure form. Since only standard laboratory techniques are involved, the overall scheme should lend itself to the preparation of a variety of 1 4 C ring- and/or chain-labeled phenols, p-quinones, and the hydroquinones derived thereform. Our main objective Tyas to find a practical synthesis for 2,3,5- and/or 2,3,6-trirnethylphenol based on readily available aliphatic starting materials. Both of these phenols represent important intermediates in the production of trimethylhydroquinone and thus of vitamin E. The basic concept comprised the construction of a suitably substituted trimethylcyclohexenone carbon (1) To whom correspondence should be addressed. (2) P. A. Wehrli and F. Pigott, Inorg. Chem., 9, 2614 (1970).
skeleton which, in a second step, could be aromatized to the corresponding trimethylphenol (Scheme I).3 Chapurlat and Dreux4 described the condensation of 3-penten-2-one with 2-butanone to yield 3,5,6trimethyl-2-cyclohexen-1-one (2). We found that the same product was obtained from the more readily accessible 4-chloro-2-pentanone5 and 2-butanone. A different substitution pattern Tyas obtained in the cyclohexenone ring when methyl vinyl ketone was condensed with 3-pentanone under strongly alkaline conditions. The uv absorption at 239 nm was compatible with the 2,3,6-trimethyl-2-cyclohexen-lone6structure, 4. The reaction we studied most carefully was the condensation of crotonaldehyde with 3-pentanone in the presence of KOH.7 Under optimal conditions a 77% yield (uv analysis) of 6 as a mixture of cis and trans isomers (nmr) vias obtained. (3) N o stereochemistry is indicated in 2, 6, and 7. (4) R . Chapurlat and J. Dreux, C. R . Acad. Sci., 253, 2361 (1961): bp for 2, 85-86O (4.5 mm). ( 5 ) A . Wohl and R . Maag, Chem. Ber., 48, 3280 (1910); 4-chloro-2pentanone is undoubtedly an intermediate in their preparation of 3-penten2-one. ( 6 ) T. Ichikawa, H. Owatari, and T. Kato, Bull. Chem. SOC.J a p . , 41, 1228 (1968): bp for 4, 92-96O (16 m m ) ; bp for 6, 92-100' (24 mm). (7) ilfter this work was completed, me became aware of a Dutch Patent abstract (6,903,484, Sept 10, 1969): bp for 6, 74-76' (12 mm).
